We investigated the influence of CO2-induced acid-base changes on the potencies of the monoquaternary relaxants (rocuronium, vecuronium, and d-tubocurarine) and bisquaternary relaxants (pancuronium, pipecuronium, and metocurine), and on the antagonism of their neuromuscular block by neostigmine. Phrenic nerve-hemidiaphragm preparations of rats were used. The pH changes were induced by changing the CO2 concentration aerating the modified Krebs solution. The potencies of the monoquaternary relaxants increased at 9% CO2 (pH 7.2) and decreased at 2.5% CO2 (pH 7.6) from those of 5% CO2 (pH 7.4) both with and without neostigmine (P < 0.05), whereas the potencies of the bisquaternary drugs did not change significantly at different pH levels from control (pH 7.4) both with and without neostigmine. The slopes of the log concentration-percent response curves of each drug were not significantly different at each pH level. The ratios of inhibitory concentrations, 50% (IC50) values with and without neostigmine at each pH value for each drug were not significantly different indicating that the neostigmine-induced antagonism for each drug was not affected by the CO2-induced acid-base changes. But the ratios of the IC50 values of the steroidal relaxants (rocuronium, vecuronium, pancuronium, and pipecuronium) were significantly lower (P < 0.05) than those of the isoquinolinium drugs (d-tubocurarine and metocurine) at each pH level, suggesting that the antagonism is enhanced at each pH level for the isoquinolinium relaxants. The difference was independent of their monoquaternary or bisquaternary nature. These results suggest that CO2 increases the potency of the monoquaternary relaxants but does not affect the bisquaternary relaxants.(ABSTRACT TRUNCATED AT 250 WORDS)